2007-01-06 02:28:40

by Chris Wright

[permalink] [raw]
Subject: [patch 15/50] IB/srp: Fix FMR mapping for 32-bit kernels and addresses above 4G

-stable review patch. If anyone has any objections, please let us know.
------------------

From: Roland Dreier <[email protected]>

struct srp_device.fmr_page_mask was unsigned long, which means that
the top part of addresses above 4G was being chopped off on 32-bit
architectures. Of course nothing good happens when data from SRP
targets is DMAed to the wrong place.

Fix this by changing fmr_page_mask to u64, to match the addresses
actually used by IB devices.

Thanks to Brian Cain <[email protected]> and David McMillen
<[email protected]> for help diagnosing the bug and testing
the fix.

Signed-off-by: Roland Dreier <[email protected]>
Signed-off-by: Chris Wright <[email protected]>
---
I just asked Linus to pull this. It fixes nasty corruption/crash
problems on 32-bit systems with > 4G of memory.

drivers/infiniband/ulp/srp/ib_srp.c | 2 +-
drivers/infiniband/ulp/srp/ib_srp.h | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)

--- linux-2.6.19.1.orig/drivers/infiniband/ulp/srp/ib_srp.c
+++ linux-2.6.19.1/drivers/infiniband/ulp/srp/ib_srp.c
@@ -1879,7 +1879,7 @@ static void srp_add_one(struct ib_device
*/
srp_dev->fmr_page_shift = max(9, ffs(dev_attr->page_size_cap) - 1);
srp_dev->fmr_page_size = 1 << srp_dev->fmr_page_shift;
- srp_dev->fmr_page_mask = ~((unsigned long) srp_dev->fmr_page_size - 1);
+ srp_dev->fmr_page_mask = ~((u64) srp_dev->fmr_page_size - 1);

INIT_LIST_HEAD(&srp_dev->dev_list);

--- linux-2.6.19.1.orig/drivers/infiniband/ulp/srp/ib_srp.h
+++ linux-2.6.19.1/drivers/infiniband/ulp/srp/ib_srp.h
@@ -87,7 +87,7 @@ struct srp_device {
struct ib_fmr_pool *fmr_pool;
int fmr_page_shift;
int fmr_page_size;
- unsigned long fmr_page_mask;
+ u64 fmr_page_mask;
};

struct srp_host {

--